1) The document compares flash and glow ATP assays using a Thermo Scientific Varioskan luminometer. 2) Flash ATP reactions produce a very fast, transient signal while glow ATP reactions produce a stable signal over longer time periods. 3) Sensitivity tests show the flash ATP assay has a lower detection limit of around 1.5 amol/well ATP using 384-well plates, around 50 times more sensitive than the glow ATP assay. Plate density affects assay sensitivity differently for flash and glow chemistries.

1. Application
Note:
AP-MIB-VARIO12-0108
Comparison of Flash and Glow ATP
Assays with Thermo Scientific Varioskan
Flash Luminometry
Jorma Lampinen, Marika Raitio, Arto Perälä, Vuokko Kytöniemi and Reija-Riitta Harinen
Thermo Fisher Scientific Oy, Vantaa, Finland
Introduction
One of the most common lumi-
nescent reactions used in research
is an ATP reaction catalysed by
firefly luciferases. In this reaction
ATP is degraded to AMP and the
energy is released as visible light.
ATP reaction can be used for the
measurement of ATP (1-3), its deg-
radation products ADP and AMP
(4) or any enzyme activity that
can be linked to ATP production
or consumption (5). It can also be
used for the measurement of firefly
enzyme activity in applications
where firefly luciferase gene is used
as a reporter gene (6-9).
Light production reaction cata-
lyzed by firefly is an oxidation
reaction where a special substrate,
D-luciferin, is oxidized and ATP is
reduced to AMP:
Luciferase
ATP + D-luciferin + O2 ------------>
Mg2+
Oxyluciferin + AMP + PPi + CO2
+ light (λmax 562 nm (10-11))
When this reaction is used to mea-
sure ATP, all other substrates are
kept in saturated concentrations,
therefore ATP is the limiting com-
ponent of the reaction and light
production is directly proportion-
al to the concentration of ATP.
Firefly-ATP reaction can be
controlled two different ways
in common applications. Light
production can be either fast flash
type reaction or a stable glow
type reaction, depending on the
assay environment. Natural wild
type firefly-ATP reaction repre-
sents this fast flash reaction and
that kind of light production is
obtained when just substrates and
luciferase are added in the reac-
tion mixture (1-2). However, this
reaction type is not very conve-
nient for applications and there-
fore the reaction is quite often
chemically stabilized to produce
stable long term light emission
(12-13). Most of the commercial
ATP assay kits use this stabilized
ATP glow reaction but lately
certain kits using ATP flash reac-
tion have been introduced. This
flash reaction offers improved
sensitivity in ATP measurement
but has a disadvantage of requir-
ing automatic dispensers for the
substrate addition. In this applica-
tion note, the basic performance
of Varioskan Flash microplate
luminometer was tested with both
glow and flash type ATP assay.
Materials and Experimental
protocols
Kits and reagents
Two luminometric ATP kits were
chosen for the evaluation: ATP
Biomass Kit HS (BioThema AB,
Sweden) representing glow type
ATP reaction, and CheckLiteTM
HS Set (Kikkoman Corp. Japan)
that is based on flash type ATP
reaction. Both kits were used
according to manufacturer’s
instructions. All tests were per-
formed with commercial
10 umol/L ATP standard
(BioThema AB, Sweden).
Instrumentation
All measurements were performed
with Thermo Scientific Varioskan
Flash multitechnology microplate
readers that were equipped with
Varioskan LumiSens high sensitivi-
ty luminometric module and auto-
matic dispenser. Instruments were
controlled using Thermo Scientific
SkanIt software.
Plastics
All measurements were per-
formed on white 96- or 384
well microplates and ATP glow
reaction was measured also
with white 1536-well plates.
The 96-well plates used were
Microlite1 96-well 12-strip plates,
384-well plates were Microlite1+
384 square well plates (both
Thermo Scientific.) and 1536-well
plates were Lumitrac 600 HiBase
1536-well plates from Greiner
Bio-One.
ATP Reaction Kinetics
ATP reaction kinetics of both flash
and glow ATP light emission reac-
tions were measured with white
384 well plates. A 30 ul aliquot of
0.01 µM ATP standard was added
into the well and plate was placed
into the microplate luminometer.
An equal volume of either flash
or glow ATP reagent was added
with the automatic dispenser and
luminescence signal was imme-
diately measured with 1 second
integration time for 10 min (flash

2. type ATP reaction) or with 500 ms
integration time for 60 min (glow
type ATP reaction). For the kinetic
measurements with ATP flash
chemistry, Varioskan Flash was
programmed as follows:
• STEP 1: Well loop step,
Execution type “by well”, Well
count 1.
Sets the instrument to perform the
following actions well by well. All
following steps are executed for
one well, then next well and so on
until all wells have been measured.
• STEP 2 (inside well loop):
Kinetic loop step “KineticLoop1”,
Readings 600, Interval 0s.
This sets the instrument to per-
form kinetic assay with 600
kinetic points. Kinetic reading is
performed as fast as possible with-
out any interval time between the
readings.
• STEP 3 (inside kinetic loop
1): Dispensing step, Dispenser
1, Dispensing volume 30 µl (for
384-well plates), Dispensing speed
“High”, Dispensing position
“Luminometric 1”, Tip priming
“No”, Dispense at reading 1.
Dispenses 30 µl of luciferin-
luciferase reagent into the well
with high speed. Dispensing is
performed simultaneously with the
first kinetic reading point.
• STEP 4 (inside kinetic loop 1):
Measurement step, Luminometric
measurement, Luminometric
optics “Normal”, Dynamic range
“AutoRange”, Measurement time
1000 ms.
Performs kinetic measurement with
1000 ms signal integration time,
sampling speed 60 points/ minute.
Total kinetic measurement time is
10 min.
Step structure of the protocol is
shown in Figure 1.
When glow type ATP chemistry
was used, a different measure-
ment protocol was used, the pro-
tocol included following actions:
• STEP 1: Kinetic loop step
“KineticLoop1”, Readings 60,
Interval 1 min.
This sets the instrument to per-
form kinetic assay with 60 kinetic
points with the interval time of
one minute between sequential
readings of the same well.
• STEP 2: (inside kinetic loop
1): Dispensing step, Dispenser
1, Dispensing volume 30 µl (for
384-well plates), Dispensing speed
“High”, Dispensing position
“Luminometric 1”, Tip priming
“No”, Dispense at reading 1.
Dispenses 30 µl of luciferin-
luciferase reagent into all sample
wells in the plate with high speed.
The first kinetic measurement is
taken simultaneously with the
dispensing.
• STEP 3: (inside kinetic loop 1):
Measurement step, Luminometric
measurement, Luminometric
optics “Normal”, Dynamic range
“AutoRange”, Measurement time
500 ms.
Performs kinetic measurement
with 500 ms integration time for
all sample wells, then waits until
defined one minute interval is
filled and then repeats the mea-
surement for 60 rounds.
Step structure of the protocol is
shown in Figure 2.
ATP Sensitivity Tests
ATP sensitivity was determined
for a set of Varioskan Flash units.
Number of Varioskan Flash units
used in each test is shown in
Table I. An ATP standard dilu-
tion series was prepared between
0.1 pM to 1 µM with 10-fold
dilutions. With 384-well plates
30 µl of each ATP dilution or
blank solution was pipetted into
the microplate wells and 30 µl of
luciferin-luciferase reagent from
either ATP flash or glow kit was
added with the instrument’s dis-
penser. For 96-well plates 50 µl
of ATP dilution or blank solution
and 50 µl of luciferin-luciferase
reagent and for 1536-well plates
5 µl of both ATP dilution and
reagent was used. Test included
with all plate types 16 replicates
of the blanks and eight replicates
of the ATP dilutions. With flash
type ATP reaction luminescence
of each well was measured after
two second delay time with
four second signal integration.
Measurement protocol used in
this assay had following proce-
dural steps:
• STEP 1: Well loop step,
Execution type “by well”, Well
count 1.
Sets the instrument to perform
the following actions well by
well. All following steps are
executed for one well, then next
well and so on until all wells have
been measured.
• STEP 2: Dispensing step,
Dispenser 1, Dispensing vol-
ume 30 µl (for 384-well
plates), Dispensing speed
“High”, Dispensing position
“Luminometric 1”, Tip priming
“No”.
Dispenses 30 µl of luciferin-
luciferase reagent into one sample
well with high speed.
• STEP 3: Measurement step,
Luminometric measurement,
Luminometric optics “Normal”,
Dynamic range “AutoRange”,
Measurement time 4000 ms, Lag
time 2 s.
This step first waits for 2 s delay
time until the signal has been
generated and then it measures
the luminescent signal for 4000
ms. Resulting luminescence result
is calculated and given for 1 sec-
ond integration time.
Step structure of the protocol is
shown in Figure 3.
Part of Thermo Fisher Scientific
Figure 1. Protocol steps for the kinetic measure-
ment of flash ATP reaction.
Figure 2. Protocol steps for the kinetic meas-
urement of glow ATP reaction.

3. Part of Thermo Fisher Scientific
When glow type ATP reaction
was measured, all wells were
first dispensed with the luciferin-
luciferase reagent and signal was
measured with one second inte-
gration time from the wells after
two minute waiting period.
• STEP 1: Dispensing step,
Dispenser 1, Dispensing vol-
ume 30 µl (for 384-well
plates), Dispensing speed
“High”, Dispensing position
“Luminometric 1”, Tip priming
“No”.
Dispenses 30 µl of luciferin-
luciferase reagent into all sample
wells in the plate with high
speed.
• STEP 2: Measurement step,
Luminometric measurement,
Luminometric optics “Normal”,
Dynamic range “AutoRange”,
Measurement time 1000 ms, Lag
time 2 min.
Step will wait two minutes at the
beginning of the step, then all
wells are measured with 1000 ms
integration time.
Step structure of the protocol is
shown in Figure 4.
Theoretical assay sensitivity was
calculated as a crossing point
between the blank subtracted
ATP calibration curve and 2*SD
level of assay blank wells and
Z-prime values were calculated
according to standard formulas
(14).
Results and Discussion
ATP Reaction Kinetics
As expected, flash and glow type
luminescence from the ATP reac-
tion show totally different kinetic
behaviour. Typical kinetic lumines-
cence curves are shown in Figure 5.
Flash type ATP luminescence reac-
tion shown in Figure 5A has very
fast signal generation where peak
maximum is reached in about four
seconds after addition of luciferin-
luciferase reagent and the signal
decays quite rapidly, showing little
below two minutes half-life time.
Therefore, this assay was measured
by integrating a signal from two
to six seconds, where maximal
signal is emitted by the reaction.
Because of this fast signal increase
and decrease, this type of ATP reac-
tion can be measured only with the
instrument equipped with auto-
matic dispenser module. If luciferin-
luciferase reagent would be added
into the ATP sample manually
outside the instrument, the signal
maximum would be totally lost
before any measurements could be
done. In addition, time difference
between dispensing the luciferin-
luciferase reagent and measuring
the signal has to be constant with
rather high precision for each well,
otherwise incorrect results would
be obtained because of signal
instability.
Glow type ATP luminescence reac-
tion in Figure 5B on the other
hand has totally different reaction
kinetics. Luminescence maximum
is reached within the minute after
ATP and luciferin-luciferase are
mixed but reaction stays stable for
much longer time period. After an
hour, there is still more than 70%
of the signal present. Therefore,
when this glow reaction is used to
measure ATP, it is possible to add
all components manually and read
the plate afterwards. This will not
cause remarkable variations in the
result values.
Figure 3. Protocol steps used to measure ATP
sensitivity assay with flash chemistry.
Figure 4. Protocol steps used to measure ATP
sensitivity assay with glow chemistry.
Figure 5. Kinetic curves of two luminescent ATP reactions. A) ATP flash reaction. B) ATP glow reaction.

4. Part of Thermo Fisher Scientific
ATP Sensitivity Tests
Sensitivity of these two ATP assays
was determined with different plate
formats, glow type ATP sensitiv-
ity with all three common plate
formats: 96-, 384- and 1536-well
plates and flash type assay only
with 96- and 384-well plates.
Results of these tests are collected
in Table I and examples of typical
ATP standard curves are shown in
Figure 6. Varioskan Flash instru-
ment is designed to have optimal
performance with 384-well plates
and that is seen from these results.
Both ATP assays using either flash
or glow type ATP chemistry show
the best sensitivity with 384-well
plates. The detection limit of ATP
with flash type chemistry and with
this plate format is about 1.5 amol/
well and with glow type chemistry
about 80 amol/well. This around
50 times sensitivity difference is
very understandable because of the
difference between the two chemis-
tries. With flash type ATP reaction,
the light from the chemical reaction
is released within the short period
of time, therefore average signal
over time is much higher. This is
seen also from the Figure 6, the
same amount of ATP in the well
produces about 30 times more light
in flash reaction. For example ATP
sample of 300 000 amol/well gives
about 5 000 000 RLU in flash
reaction when glow reaction gives
only 150 000 RLU with the same
amount of ATP.
When the plate format is changed
from 384-wells to 96-wells, assay
sensitivity is decreased. This effect
is not remarkable with flash type ATP
assay, only about 20% decrease in the
detection limit, but with glow type
ATP assay the sensitivity is decreased
almost 10 times. Therefore the dif-
ference between flash and glow assay
sensitivity with 96-well plates is much
bigger with 96-well plates than with
384-well plates, being about 400
times.
The very high density 1536-well
plate was tested only with
glow type ATP chemistry and
it was shown to produce about
the same ATP sensitivity than
96-well plate, having the detec-
tion limit value of 200 amol/well
ATP, both slightly lower than
with 384-well plate.
Plate
format
ATP Flash chemistry ATP Glow Chemistry
Glow/
Flash
ratio
Average detection
limit (amol/well))
Range, (Min-Max) Number of Varioskan
Flash units
Average detection
limit (amol/well))
Range, (Min-
Max)
Number of Varioskan
Flash units
96 1.8 0.8 – 2.6 7 130 600 - 900 4 409
384 1.5 0.5 – 2.9 30 80 60 - 130 6 52
1536 -- -- -- 200 180 - 210 2 --
Figure 6. Typical ATP calibration curves with different plate formats. A) Flash type ATP assay. B)
Glow type ATP assay.
Table I. Sensitivity test results of flash and glow ATP assays with Varioskan Flash.

5. Part of Thermo Fisher Scientific
All sensitivity tests were performed
with several Varioskan Flash units.
As expected, there are differences
between the performances of the
individual instruments in these
ATP assays. Every instrument is
always individual in its perfor-
mance because there are always
some differences in electronics,
noise levels etc. The total perfor-
mance range in the detection limits
between the instruments is any-
how amazingly narrow. When the
average ATP detection limit with
the flash chemistry and 384-well
plates is for example 1.8 amol/
well, all tested 30 instrument units
had their detection limit below 3
amol/well ATP. Figure 7. shows
the cumulative distribution of
these detection limit values of the
individual Varioskan Flash instru-
ments. This distribution is clearly
following expected cumulative
Gaussian distribution.
Another very important assay
quality parameter is Z-prime
value. It is a numerical calculation
that reflects the assay capability
to separate positive results from
the negative ones (14). Theoretical
maximum of the Z-prime value is
1.0 that suggests that reliability
of the separation of two signals
is 100.0%. This is of course
never possible to be met in real
life because there is always some
standard deviation in the signal
values. If the Z-prime value is over
0.5, that assay can be normally
considered good and reliable (14)
but sometimes values around
0.6 or higher might be required.
Calculated Z-prime values against
the assay blanks over the whole
ATP calibration series are shown
for both flash and glow ATP
chemistries and all plate formats in
Figure 8.
Figure 8. clearly shows the excel-
lent reliability of these ATP assays.
With the flash type ATP assay,
that 3 amol/well ATP sample
that is very close to the theoreti-
cal detection limit shows already
almost sufficient Z-prime values.
Theoretical detection limit is
calculated based on blank SD
only when Z-prime calculation
formula includes both blank and
standard SD values. Therefore, cal-
culated theoretical detection limit
is always below the concentration
range where good Z-prime values
are obtained. With the ATP sample
of 30 amol/well Z-prime values
are clearly over required 0.5
level, with both 96- and 384-well
formats.
When the Z-prime values of
glow type ATP assay are ana-
lyzed, it is immediately noticed
that higher ATP amounts than
with flash chemistry are required
for the good z-prime level. About
300 amol/well ATP is required
before calculated Z-prime val-
ues are constantly over required
level.
Figure 7. Cumulative distribution of the Varioskan Flash detection limits in flash type ATP assay
with 384-well plates.
Figure 8. The Z-prime values of both flash and glow type ATP assays measured with 96-, 384- and
1536-well plate formats.

6. Part of Thermo Fisher Scientific
Conclusions
These results clearly show the high
sensitivity and reliability of ATP
assays with Varioskan Flash lumi-
nometry. ATP assay using flash type
chemistry gives better assay sensitivi-
ty than assays using glow type chem-
istry and 384-well plate format gives
the best sensitivity with both types
of chemistries. For the statistically
reliable results, around 10-30 amol
of ATP is required in the sample
with flash type ATP assay and about
300 -500 amol/well is required for
the glow type assay.
References
1. Lundin, A., et al. 1976. Continuous
monitoring of ATP-converting reactions by
purified firefly luciferase. Anal. Biochem.
75, pp. 611 - 620.
2. Webster, J.J. and Leach, F. 1980.
Optimization of the firefly luciferase assay
for ATP. J. Applied Biochem. 2, pp. 469 -
79.
3. Karl, D.M. 1980. Cellular nucleotide
measurements and applications in micro-
bial ecology. Microbiol. Rev. 44, pp. 739
- 96.
4. Lundin, A. et al. 1986. Estimation
of biomass in growing cell lines by
adenosine triphosphate assay. Methods in
Enzymology 133, pp. 27 - 42.
5. Fan, F. and Wood K.V. 2007.
Bioluminescent Assays for High-
Throughput Screening. Assay and Drug
Development Technologies. 5. pp.
127-136.
6. Ow, D.W. et al. 1986. Transient and
stable expression of the firefly luciferase
gene in plant cells and transgenic plants.
Science 234, pp. 856 - 859.
7. De Wet, J.R. et al. 1987. Firefly
luciferase gene: Structure and expression in
mammalian cells. Mol. Cell. Biol. 7, 725 -
737.
8. Brasier, A.R. et al. 1989. Optimized use
of firefly luciferase assay as a reporter gene
in mammalian cell lines. BioTechniques. 7,
pp. 1116 - 1122.
9. Koncz, C. et al. 1990. Bacterial and
Firefly Luciferase Genes in Transgenic
Plants: Advantages and Disadvantages of
a Reporter Gene. Dev. Genet. 11, pp. 224
- 232.
10. Seliger, H.H. et al. 1964. The Spectral
Distribution of Firefly Light. J. Gen. Physiol.
48, pp. 95-104.
11. Ugarova, N.N. and L.Y. Brovko. 2002.
Protein structure and bioluminescent spectra
for firefly bioluminescence. Luminescence
17, pp. 321-330.
12. Lundin, et al. 1982. Optimized biolu-
minescence assay of creatine kinase and cre-
atine kinase B-subunit activity. Clin. Chem.
28, pp. 609-614.
13. Wood, K V. et al. 1991. Recent
Advances and Prospects for Use of Beetle
Luciferases as Genetic Reporters. In:
Bioluminescence & Chemiluminescence:
Current Status, Proceedings of the
VIth International Symposium on
Bioluminescence and Chemiluminescence,
Stanley, P. E., et al., Editors, John Wiley &
Sons, Chichester, United Kingdom. Pp. 543-
546.
14. Carter, J.M. 2003. A Guide to Assay
Development. D&MD Publications. Pp.
12/10 – 12/11.
North America:
USA / Canada
+1 800 522 7763
Europe:
Austria
+43 1 801 40 0
Belgium
+32 2 482 30 30
France
+33 2 2803 2000
Germany national toll free
08001-536 376
Germany international
+49 6184 90 6940
Italy
+39 02 02 95059 434-254
Netherlands
+31 76 571 4440
Nordic countries
+358 9 329 100
Russia/CIS
+7 (812) 703 42 15
Spain/Portugal
+34 93 223 09 18
Switzerland
+41 44 454 12 12
UK/Ireland
+44 870 609 9203
Asia:
China
+86 21 6865 4588 or
+86 10 8419 3588
India
+91 22 6716 2200
Japan
+81 45 453 9220
Other Asian countries
+852 2885 4613
Countries not listed:
+49 6184 90 6940 or
+33 2 2803 2000
www.thermo.com/
readingroom
www.thermo.com/mpi
In addition to these offices,
Thermo Fisher Scientific
maintains a network of
representative organizations
throughout the world.
© 2008 Thermo Fisher Scientific
Inc. All rights reserved. All
trademarks are the property of
Thermo Fisher Scientific Inc. and
its subsidiaries. Specifications,
terms and pricing are subject to
change. Not all products are
available in all countries. Please
consult your local sales repre-
sentative for details.